Andrew Oehler

Cardiac rehabilitation (CR) is a multicomponent secondary prevention intervention that improves functional capacity and quality of life across cardiac diagnoses, yet remains markedly underused in heart failure (HF). Existing reviews treat efficacy, coverage, delivery, and equity separately; this narrative review integrates them to clarify why uptake in HF remains low despite an established evidence base. We searched PubMed/MEDLINE and the Cochrane Library from January 2009 through December 2025, prioritizing clinical practice guidelines, landmark trials, systematic reviews, registry analyses, and U.S. policy documents, with narrative synthesis across efficacy, coverage, delivery, and equity. In heart failure with reduced ejection fraction (HFrEF), exercise-based CR consistently improves peak oxygen consumption by roughly 15 to 17 percent, improves health status, and reduces short-term all-cause hospitalization. In heart failure with preserved ejection fraction (HFpEF), supervised exercise improves functional capacity and quality of life, but a large pragmatic outcome trial and Medicare coverage remain lacking. In appropriately selected, clinically stable patients, hybrid and virtual delivery produce comparable short-term functional and quality-of-life gains, with more limited direct evidence for hard clinical endpoints. Closing the HFpEF coverage gap, automating HFrEF referral, diversifying delivery, and embedding equity across the CR pipeline are actionable priorities for realizing CR’s potential in HF.

BACKGROUND: Coronary plaque features are imaging biomarkers of cardiovascular risk, but less is known about sex-specific patterns in their prognostic value. This study aimed to define sex differences in the coronary atherosclerotic phenotypes assessed by artificial intelligence–based quantitative computed tomography (AI-QCT) and the associated risk of major adverse cardiovascular events (MACEs). METHODS: Global multicenter registry including symptomatic patients with suspicion of coronary artery disease referred for coronary computed tomography angiography. AI-QCT analyzed 16 coronary artery disease features. The primary end point was MACE defined as death, myocardial infarction, late revascularization, cerebrovascular events, unstable angina, and congestive heart failure. RESULTS: Among 3551 patients (mean age, 59±12 years; 49.5% women), MACE occurred in 3.2% of women and 6.1% of men during an average follow-up of 4.8±2.2 years. The AI-QCT features total plaque volume, noncalcified plaque, calcified plaque, and percentage atheroma volume were significantly higher in men ( P P 3 increase in total plaque volume, relative risk increased by 17.7% (95% CI, 1.12–1.24) in women versus 5.3% (95% CI, 1.03–1.07) in men ( P _interaction P _interaction =0.0015); and for calcified plaque, relative risk increased by 22.9% (95% CI, 1.14–1.33) versus 5.4% (95% CI, 1.01–1.10; P _interaction =0.0012), respectively. Similarly, for percentage atheroma volume, the risk was higher in women. The findings remained unchanged when restricted to a secondary composite end point (death and myocardial infarction). CONCLUSIONS: The AI-QCT plaque features, total plaque volume, noncalcified plaque, calcified plaque, and percentage atheroma volume, conferred a higher relative MACE risk in women and may prompt more aggressive antiatherosclerotic therapy and reinforced preventive interventions. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT04279496.

Purpose: Cardiac rehabilitation (CR) has been proven to reduce mortality and morbidity in patients with cardiovascular disease. Machine learning (ML) techniques are increasingly used to predict healthcare outcomes in various fields of medicine including CR. This systemic review aims to perform critical appraisal of existing ML-based prognosis predictive model within CR and identify key research gaps in this area. Review Methods: A systematic literature search was conducted in Scopus, PubMed, Web of Science, and Google Scholar from the inception of each database to January 28, 2024. The data extracted included clinical features, predicted outcomes, model development, and validation as well as model performance metrics. Included studies underwent quality assessments using the IJMEDI and Prediction Model Risk of Bias Assessment Tool checklist. Summary: A total of 22 ML-based clinical models from 7 studies across multiple phases of CR were included. Most models were developed using smaller patient cohorts from 41 to 227, with one exception involving 2280 patients. The prediction objectives ranged from patient intention to initiate CR to graduate from outpatient CR along with interval physiological and psychological progression in CR. The best-performing ML models reported area under the receiver operating characteristics curve between 0.82 and 0.91, with sensitivity from 0.77 to 0.95, indicating good prediction capabilities. However, none of them underwent calibration or external validation. Most studies raised concerns about bias. Readiness of these models for implementation into practice is questionable. External validation of existing models and development of new models with robust methodology based on larger populations and targeting diverse clinical outcomes in CR are needed. KEY PERSPECTIVE: What is novel? This systematic review is the first to critically appraise existing machine learning (ML)-based clinical prognostic models in cardiac rehabilitation (CR). It identifies key research gaps in ML application in CR and provides constructive suggestions for improving the quality of ML models specifically in CR. What are the clinical and/or research implications? The review demonstrates various clinical challenges in the CR setting that ML models have been used to address and their corresponding outcomes. It reveals potential limitations in existing models and offers recommendations to enhance the readiness of current and future models for implementation in clinical settings as clinical decision-support tools.

Purpose CR has been proven to reduce mortality and morbidity in patients with CVD. ML techniques are increasingly used to predict healthcare outcomes in various fields of medicine including CR. This systemic review aims to perform critical appraisal of existing ML based prognosis predictive model within CR and identify key research gaps in this area. Review methods A systematic literature search was conducted in Scopus, PubMed, Web of Science and Google Scholar from the inception of each database to 28th January 2024. The data extracted included clinical features, predicted outcomes, model development and validation as well as model performance metrics. Included studies underwent quality assessments using the IJMEDI. Summary 22 ML-based clinical models from 7 studies across multiple phases of CR were included. Most models were developed using smaller patient cohorts from 41 to 227, with one exception involving 2280 patients. The prediction objectives ranged from patient intention to initiate CR to graduate from outpatient CR along with interval physiological and psychological response to CR. The best-performing ML models reported AUC between 0.82 and 0.91, sensitivity from 0.77 to 0.95, indicating good prediction capabilities. However, none of them underwent calibration or external validation. Most studies raised concerns for bias. Readiness of these models for implement into practice is questionable. External validation of existing models and development of new models with robust methodology based on larger populations and targeting diverse clinical overcomes in CR are needed.